What's new
What's new

Machining Thin Wall SS Component (0.004", 100 micron wall)

dgoodell

Plastic
Joined
Sep 10, 2018
Location
Cleveland, Ohio
I have a thin wall (0.004", 100 micron) 316 SS component that we are finding very difficult to get manufactured. Here's a link to the drawing.

The component is part of a prototype 1-watt stirling engine. If the design is successful, this could lead to a project to build these engines for space use to power sensor packages on various planets and moons in the solar system.

We didn't have much luck getting quotes to get this part made as-is so we decided to manage the manufacture of this part ourselves. It was easy enough to get the outer features turned and a blind a hole drilled into the middle. The trouble came we tried to plunge EDM the ID. Whenever we did the thin wall section the part would just disintegrate. It didn't seem that we could EDM something that thin.

Now we're thinking that we should encase the part in a fusible alloy (like Cerrotru) to make the wall thicker prior to EDM.

Alternatively, we could finish the ID first, then fill the part with Cerrotru, then finish the o.d. features to get the required wall thickness.

Does anyone have any idea about what approach to manufacturing this part might make the most sense?
 
Does it have to be 316? I had a colleague many years back who made very thin section things by electro depositing nickel on a form. Make the forn, coat with a conductor, deposit the nickel, dissolve the form away.
 
CAD rule #1- Don't draw it if you don't know how it can be made! IMO, the suggestions for a fabrication or deposits are good. The precision bellows are made by depositing and can be much thinner. Maybe it could be ground on a mandrel, but no matter what you do, it won't be cheap.
 
Interesting how they dimension the part on the OD and ID of the thin wall section based on the tolerances given you could make it a .01 thick wall if it weren’t for that thickness callout. If I was you I would shoot for a .0055 thick wall since the tolerance allows it on the OD. Make a mandrel and turn it. It may only be a thou and a half but when things are that thin it makes a big difference. Rough it out then with a very sharp positive angle bit cut off the last bit of material it may take a couple of passes but you can get it there. The trick is to apply as little cutting force without dragging the tool creating heat and dulling the tool which in turn creates more pressure. Since it’s 316 you should be ok as it’s very stable.
 
Does it have to be 316? I had a colleague many years back who made very thin section things by electro depositing nickel on a form. Make the forn, coat with a conductor, deposit the nickel, dissolve the form away.

Anything could work but it needs to be strong enough to hold 150 psi with one end of the tube at about 750 deg F while minimizing thermal conduction along its length. I think pure nickel has much higher thermal conductivity and lower strength than stainless steel but that doesn't mean it won't work.
 
CAD rule #1- Don't draw it if you don't know how it can be made! IMO, the suggestions for a fabrication or deposits are good. The precision bellows are made by depositing and can be much thinner. Maybe it could be ground on a mandrel, but no matter what you do, it won't be cheap.

A vendor drew this up for us on a contract. Of course, we had to get all of the components manufactured for this engine so we gave the same vendor a contract to go procure the components. They have been struggling to get this part made. They told us they are going to have a cylinder that is 0.012" thick rather than 0.004-6" thick made at a cost of several thousands each with some other geometric modifications they didn't ask us about first. That's why we're kind of looking into figuring out how to get this made ourselves.

We had a guy who knew how to make these parts for us but he signed a contract with some company and he can't make these for anyone else now.
 
The over burn of a Sinker is going to blow the walls out. An experienced guy with a Wire machine will be able to burn the ID out. (Maybe)

You have .005mm on Cylindricity call out, on a Feature that needs to be in Position to .01mm!!!! IMHO you are never going to be able to achieve that. I can't even think of a plausible way to inspect something like that. A CMM won't do it, you can't put hard gages to it, Fit and ultimately Function is the only test I can see. As an assembly, the part might be manufacturable as is.

R
 
I believe you need to go back to drawing board and come up with something that is more easily manufactured. Maybe start with hypodermic tubing and have the other 2 pieces laser welded/ brazed on? I have seen .004 laser weld before many years ago. I feel sure with modern technology something viable could be found.
 
I believe you need to go back to drawing board and come up with something that is more easily manufactured. Maybe start with hypodermic tubing and have the other 2 pieces laser welded/ brazed on? I have seen .004 laser weld before many years ago. I feel sure with modern technology something viable could be found.

Yes, hypo tubing with silver soldered fittings would be the way I would go with this. Would make it dramatically cheaper to build and much less effort. Proper soldering and cleaning would be needed to ensure no residual flux was left behind.
 
I believe you need to go back to drawing board and ...

Well, I have not even got beyond the .434 +/-.001 @ 3.5" depth bore with a flat bottom thing.

Then the 750F temperature with conduction issues... on SST?
I ain't no metallurgist, but Inco ( even if it's lower strength than SS ) is gonna be a better choice at those temperatures.
 
Well, I have not even got beyond the .434 +/-.001 @ 3.5" depth bore with a flat bottom thing.

Then the 750F temperature with conduction issues... on SST?
I ain't no metallurgist, but Inco ( even if it's lower strength than SS ) is gonna be a better choice at those temperatures.

Fair enough on calling out the temperature. A286 would work, but a Inconel 718 might be a better choice. And that means brazing, not soldering for assembly.
 
Laser welding from multiple parts would be the way. You should first find a company that are experts in laser welding and talk to them. Then redesign the part after their input.

/Staffan
 
I see round bar on a swiss lathe, turning one end features, deep drilling, turning other end features, honing ID, and a well cooled grind operation for the OD, part on a quill.
 
I second You Mechanola thats how i would make it. Once you have got that ID honed out with the right tooling approach you should be able to turn it pretty thin on a mandaral, certainly into the sub 20 thou range and then grind from there.

Yeah its going to be expensive and also very easily damaged, but design shit with zero thought to manufacture and thats what you get. Still can not realy see why everyone wants to use Stirling engines in space, thermopiles and solar are far more proven and significantly less liable to fail unlike a moving component engine.
 
I second You Mechanola thats how i would make it. Once you have got that ID honed out with the right tooling approach you should be able to turn it pretty thin on a mandaral, certainly into the sub 20 thou range and then grind from there.

Yeah its going to be expensive and also very easily damaged, but design shit with zero thought to manufacture and thats what you get. Still can not realy see why everyone wants to use Stirling engines in space, thermopiles and solar are far more proven and significantly less liable to fail unlike a moving component engine.

I absolutely agree with you about the design not considering manufacturing. Our contractor who designed this part is also in charge of getting it made and they haven't had much success. We're trying to figure this out ourselves kind of late in the game so all the feedback (even the negative) is very beneficial. It sounds like it may be best to redesign this part ourselves working with a manufacturer.

The stirling engines are attractive because they open up a lot of possibilities in space, without them there are lots of things we simply can't do. If you want enough power for human missions to mars then some kind of heat engine will be required to produce enough power. The thermoelectrics are so inefficient that can cause problems, solar power only works when you have ample sunlight. It's even difficult to make a moon lander that's solar powered because the moon's night is 15 days long.
 
Well, I have not even got beyond the .434 +/-.001 @ 3.5" depth bore with a flat bottom thing.

Then the 750F temperature with conduction issues... on SST?
I ain't no metallurgist, but Inco ( even if it's lower strength than SS ) is gonna be a better choice at those temperatures.

The 316 stainless loses a lot of its strength at that temperature but it has enough margin that it should still be ok. It meets the BPVC (boiler pressure vessel code) requirements taking into account strength loss at that temp. For space flight, we can use Inconel 718 up to 650C (1202 F) and we can use Mar-M-247 at up to around 850C (1562 F). You can get a little hotter if you accept reduced life or higher risk.
 
I absolutely agree with you about the design not considering manufacturing. Our contractor who designed this part is also in charge of getting it made and they haven't had much success. We're trying to figure this out ourselves kind of late in the game so all the feedback (even the negative) is very beneficial. It sounds like it may be best to redesign this part ourselves working with a manufacturer.

The stirling engines are attractive because they open up a lot of possibilities in space, without them there are lots of things we simply can't do. If you want enough power for human missions to mars then some kind of heat engine will be required to produce enough power. The thermoelectrics are so inefficient that can cause problems, solar power only works when you have ample sunlight. It's even difficult to make a moon lander that's solar powered because the moon's night is 15 days long.

Yeah ok, spose im not use to have to consider a 15 day night, on the plus side, 15 x24 hours would sure mean i might not wake up tired, but can't imagine sleeping that long with out a few pees even at my age!

If your haveing issue getting it made, you need to find some of the medical suppliers, theres plenty of people there in the states that could make that for you, medical type stuff are use to long thin tubes its kinda there starting point for stents and such, might be worth contacting Implex (sp) who is a member here and makes all sorts of weird bits but this might be too big for him. Im pretty dang certain i could make the tubular part here with out issue and that's just turning it, my problem is i could not hit some of the other tolerances on that part on the end bits. I just don't mesure to that kinda accuracy but i can certainly turn very very thin walled tubes Would not be entirely wrong to say i have done near half that wall thickness before in 316.
 








 
Back
Top